imzml 0.0.2 → 0.1.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 1004a80f49dadaf56d396f7baa306fd55d5708c9
-  data.tar.gz: 6188392c95e532991ebd80035013f4a8acf2e1bc
+  metadata.gz: f7be2442cc6f7ce7a1503a2a7c5380c9509043b2
+  data.tar.gz: f76e6a4ff83c1335b90389d6ea6348a2df7a3656
 SHA512:
-  metadata.gz: e5fee636296de8d17e7f8bf927bdbb351288277ccea8919992cb6cdeea4134a58ecb81917b85aeba4b546321f0cace042e2474bfee9f8829a2827488d01e8289
-  data.tar.gz: 6ac5981b89e67152ec0277de0cc6aa133965272799fa5ddc9c7fcf2c9a0524b19176984ee9be19d602aca50ef1bfd3f9d2b51e13c28d528bcd9a5ce7d8cd9f7b
+  metadata.gz: b6fa51fac3db3e0ed575eca4152b184063ab248da7c91989d1726781919119cb6def12d31965d7cea8a1b424d72d66f93da5ee3077b17ef02cc7dd837bc73409
+  data.tar.gz: d56ab3b35465d58974c80d3e7dd8f1037bd13be794261765ffcee13194510d893ea8e684cffd4a06a82614a954c5c74d7c3f79489b8447fde8ba123f9ea4f1b3

data/lib/core_ext/array.rb

@@ -0,0 +1,23 @@
+class Array
+  
+  def binary_search(value, first = true)
+    
+    if (self.size > 2)
+      middle_index = self.size/2
+      middle = self[middle_index]
+
+      if (middle > value)
+        self[0..middle_index].binary_search(value, first)
+      else
+        self[middle_index..self.size].binary_search(value, first)
+      end
+    else
+      if first
+        self.first
+      else
+        self.last
+      end
+    end
+  end
+  
+end

data/lib/imzml/metadata/run/spectrum.rb

@@ -4,6 +4,14 @@ module ImzML
     
     class BinaryData
       
+      # Binary data types, always little endian
+      BINARY_TYPE_8BIT_INTEGER = "IMS:1100000"
+      BINARY_TYPE_16BIT_INTEGER = "IMS:1100001"
+      BINARY_TYPE_32BIT_INTEGER = "MS:1000519"
+      BINARY_TYPE_64BIT_INTEGER = "MS:1000522"
+      BINARY_TYPE_32BIT_FLOAT = "MS:1000521"
+      BINARY_TYPE_64BIT_FLOAT = "MS:1000523"
+      
       # A data array of m/z values
       MZ_ARRAY = "MS:1000514"
       
@@ -22,11 +30,48 @@ module ImzML
       attr_accessor :encoded_length
       EXTERNAL_ENCODED_LENGHT = "IMS:1000104"
       
+      # Path to the external binary file
+      attr_accessor :filepath
+      
+      # Binary values type [:int8, :int16, :int32, :int64, :float32, :float64]
+      attr_accessor :type
+      
       # grabs the actual binary data from disk
-      def data
+      def data(cached = true)
+        
+        # Return the data from the cache
+        return @cached_data if cached && !@cached_data.nil?
+        
+        # Remove possible data from the cache
+        @cached_data = nil
         
+        # Switch binary pattern reading type
+        pattern = case type
+        when :int8
+          "C"
+        when :int16
+          "S"
+        when :int32
+          "L"
+        when :int64
+          "Q"
+        when :float32
+          "e"
+        when :float64
+          "E"
+        end
+
+        # Read data based on metadata
+        data = IO.binread(@filepath, @encoded_length.to_i, @offset.to_i).unpack("#{pattern}*")
+        
+        # Save data only if user want's to cache it, saving take some CPU
+        @cached_data = data if cached
       end
       
+      private
+      
+      attr_accessor :cached_data
+      
     end
     
     # Attributes to describe the position of a spectrum in the image.
@@ -45,6 +90,36 @@ module ImzML
     #
     # Represented by class BinaryData
     attr_accessor :intensity_binary
+    
+    def intensity(at, interval)
+      
+      # read whole the binary data
+      mz_array = mz_binary.data
+      intensity_array = intensity_binary.data
+      
+      default_from, default_to = mz_array.first, mz_array.first
+      
+      # find designated intensity
+      if !at
+        from = default_from
+        to = default_to
+      else
+        from = at - interval
+        from = default_from if from < 0
+        to = at + interval
+        to = default_to if to > mz_array.last
+      end
+      
+      # find values in mz array
+      low_value = mz_array.binary_search(from, false)
+      low_index = mz_array.index(low_value)
+      high_value = mz_array.binary_search(to)
+      high_index = mz_array.index(high_value)
+  
+      # sum all values in subarray
+      intensity_array[low_index..high_index].inject{|sum, x| sum + x}
+      
+    end
 
     
   end

data/lib/imzml/metadata.rb

@@ -19,18 +19,6 @@ module ImzML
     
     # All mass spectra and the acquisitions underlying them are described and attached here
     attr_accessor :spectrums
-        
-    # Binary data types, always little endian
-    BINARY_TYPE_8BIT_INTEGER = "IMS:1100000"
-    BINARY_TYPE_16BIT_INTEGER = "IMS:1100001"
-    BINARY_TYPE_32BIT_INTEGER = "MS:1000519"
-    BINARY_TYPE_64BIT_INTEGER = "MS:1000522"
-    BINARY_TYPE_32BIT_FLOAT = "MS:1000521"
-    BINARY_TYPE_64BIT_FLOAT = "MS:1000523"
-    
-    # both can have one of the symbols [:int8, :int16, :int32, :int63, :float32, :float64]
-    attr_accessor :mz_binary_data_type
-    attr_accessor :intensity_binary_data_type
 
   end
 end

data/lib/imzml/parser.rb

@@ -6,12 +6,22 @@ module ImzML
 
     attr_reader :metadata
 
-    def initialize(filepath)
+    def initialize(filepath, binary_filepath = nil)
 
       sax = ImzML::Sax.new
+      
+      # If no external file were specified, behave like the file has the same name
+      # as the metadata
+      if binary_filepath.nil?
+        name = filepath.split(".")[0..-2].join(".")
+        binary_filepath = [name, "ibd"].join(".")
+      end
+      sax.binary_filepath = binary_filepath      
+
+      # parse the XML
       Ox.sax_parse(sax, File.open(filepath))
       @metadata = sax.metadata
-
+      
     end
 
   end
@@ -19,6 +29,11 @@ module ImzML
   class Sax < ::Ox::Sax
 
     attr_reader :metadata
+    attr_accessor :binary_filepath
+    
+    # Both can have one of the symbols [:int8, :int16, :int32, :int64, :float32, :float64]
+    attr_accessor :mz_binary_data_type
+    attr_accessor :intensity_binary_data_type
 
     def initialize()
       @metadata = Metadata.new
@@ -224,22 +239,7 @@ module ImzML
             point.y = element[:value].to_i
           end
         end
-
-        # [
-        #   {:cvRef=>"IMS", :accession=>"IMS:1000401", :name=>"top down", :value=>""},
-        #   {:cvRef=>"IMS", :accession=>"IMS:1000413", :name=>"flyback", :value=>""},
-        #   {:cvRef=>"IMS", :accession=>"IMS:1000480", :name=>"horizontal line scan", :value=>""},
-        #   {:cvRef=>"IMS", :accession=>"IMS:1000491", :name=>"linescan left right", :value=>""},
-        #   {:cvRef=>"IMS", :accession=>"IMS:1000042", :name=>"max count of pixel x", :value=>"3"},
-        #   {:cvRef=>"IMS", :accession=>"IMS:1000043", :name=>"max count of pixel y", :value=>"3"},
-        #   {:cvRef=>"IMS", :accession=>"IMS:1000044", :name=>"max dimension x", :value=>"300", :unitCvRef=>"UO", :unitAccession=>"UO:0000017", :unitName=>"micrometer"},
-        #   {:cvRef=>"IMS", :accession=>"IMS:1000045", :name=>"max dimension y", :value=>"300", :unitCvRef=>"UO", :unitAccession=>"UO:0000017", :unitName=>"micrometer"},
-        #   {:cvRef=>"IMS", :accession=>"IMS:1000046", :name=>"pixel size x", :value=>"100", :unitCvRef=>"UO", :unitAccession=>"UO:0000017", :unitName=>"micrometer"},
-        #   {:cvRef=>"IMS", :accession=>"IMS:1000047", :name=>"pixel size y", :value=>"100", :unitCvRef=>"UO", :unitAccession=>"UO:0000017", :unitName=>"micrometer"},
-        #   {:cvRef=>"MS", :accession=>"MS:1000836", :name=>"dried dropplet", :value=>""},
-        #   {:cvRef=>"MS", :accession=>"MS:1000835", :name=>"matrix solution concentration", :value=>"10"},
-        #   {:cvRef=>"MS", :accession=>"MS:1000834", :name=>"matrix solution", :value=>"DHB"}
-        # ]
+        
       end
 
       # parse processing methods
@@ -285,23 +285,23 @@ module ImzML
         number_type = nil
         group.each do |param|
           number_type = case param[:accession]
-          when Metadata::BINARY_TYPE_8BIT_INTEGER
+          when Spectrum::BinaryData::BINARY_TYPE_8BIT_INTEGER
             :int8
-          when Metadata::BINARY_TYPE_16BIT_INTEGER
+          when Spectrum::BinaryData::BINARY_TYPE_16BIT_INTEGER
             :int16
-          when Metadata::BINARY_TYPE_32BIT_INTEGER
+          when Spectrum::BinaryData::BINARY_TYPE_32BIT_INTEGER
             :int32
-          when Metadata::BINARY_TYPE_64BIT_INTEGER
+          when Spectrum::BinaryData::BINARY_TYPE_64BIT_INTEGER
             :int64
-          when Metadata::BINARY_TYPE_32BIT_FLOAT
+          when Spectrum::BinaryData::BINARY_TYPE_32BIT_FLOAT
             :float32
-          when Metadata::BINARY_TYPE_64BIT_FLOAT
+          when Spectrum::BinaryData::BINARY_TYPE_64BIT_FLOAT
             :float64
           end
           
           break if !number_type.nil?
         end
-        @metadata.send("#{@binary_type.to_s}_data_type=", number_type) if !number_type.nil?
+        self.send("#{@binary_type.to_s}_data_type=", number_type) if !number_type.nil?
       end
       
       # save info about binary
@@ -310,6 +310,8 @@ module ImzML
         
         # convert chosen type to mz_binary/intensity_binary property selector
         binary_data = spectrum.send(@binary_type.to_s)
+        binary_data.filepath = binary_filepath
+        binary_data.type = self.send("#{@binary_type}_data_type")
         case element[:accession]
         when ImzML::Spectrum::BinaryData::EXTERNAL_ARRAY_LENGTH
           binary_data.length = element[:value].to_i

data/lib/imzml/version.rb

@@ -1,5 +1,5 @@
 module ImzML
   
-  VERSION = "0.0.2"
+  VERSION = "0.1.0"
   
 end

data/lib/imzml.rb

@@ -4,6 +4,7 @@ require "obo_ext/parser"
 require "obo_ext/stanza"
 
 require "core_ext/string"
+require "core_ext/array"
 
 require "imzml/obo"
 require "imzml/metadata"

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: imzml
 version: !ruby/object:Gem::Version
-  version: 0.0.2
+  version: 0.1.0
 platform: ruby
 authors:
 - Ondra Beneš
@@ -73,6 +73,7 @@ files:
 - data/imagingMS.obo
 - data/psi-ms.obo
 - data/unit.obo
+- lib/core_ext/array.rb
 - lib/core_ext/string.rb
 - lib/imzml.rb
 - lib/imzml/metadata.rb
@@ -83,7 +84,6 @@ files:
 - lib/imzml/metadata/run/spectrum.rb
 - lib/imzml/metadata/scan_settings.rb
 - lib/imzml/metadata/scan_settings/image.rb
-- lib/imzml/metadata/spectrum.rb
 - lib/imzml/obo.rb
 - lib/imzml/parser.rb
 - lib/imzml/version.rb

data/lib/imzml/metadata/spectrum.rb

@@ -1,79 +0,0 @@
-module ImzML
-
-  class Spectrum
-
-    attr_accessor :id
-    attr_accessor :mz_array_external_offset
-    attr_accessor :mz_array_external_encoded_length
-    attr_accessor :intensity_array_external_offset
-    attr_accessor :intensity_array_external_encoded_length
-
-    def intensity(data_path, at, interval)
-
-      raise "Interval cannot be nil" if !interval
-
-      # read array and intensity data
-      mz_array = mz_array(data_path)
-      intensity_array = intensity_array(data_path)
-
-      default_from, default_to = mz_array.first, mz_array.first
-
-      # find designated intensity
-      if !at
-        from = default_from
-        to = default_to
-      else
-        from = at - interval
-        from = default_from if from < 0
-        to = at + interval
-        to = default_to if to > mz_array.last
-      end
-
-      # find values in mz array
-      low_value = search_binary(mz_array, from)
-      low_index = mz_array.index(low_value)
-      high_value = search_binary(mz_array, to)
-      high_index = mz_array.index(high_value)
-
-      # sum all values in subarray
-      intensity_array[low_index..high_index].inject{|sum, x| sum + x}
-    end
-
-    def mz_array(data_path)
-      IO.binread(data_path, @mz_array_external_encoded_length.to_i, @mz_array_external_offset.to_i).unpack("e*")
-    end
-
-    def intensity_array(data_path)
-      IO.binread(data_path, @intensity_array_external_encoded_length.to_i, @intensity_array_external_offset.to_i).unpack("e*")
-    end
-
-    private
-
-    def search_binary(array, value, first = true)
-
-      if (array.size > 2)
-        middle_index = array.size/2
-        middle = array[middle_index]
-
-        if (middle > value)
-          search_binary(array[0..middle_index], value, first)
-        else
-          search_binary(array[middle_index..array.size], value, first)
-        end
-      else
-        if first
-          array.first
-        else
-          array.last
-        end
-      end
-
-    end
-
-    def search_last(array, value)
-
-    end
-
-  end
-
-end